Room: Karl Dean Ballroom C
Purpose: Current technologies for displaying and manipulating volumetric medical imaging are limited to presenting this 3D information on a 2D computer display. Providers must mentally reconstruct 3D features after serially viewing 2D slices. This can lead to increased cognitive workload and difficulty in identifying non-planar anatomic features, and inaccuracy in contouring structures. We present a novel virtual reality (VR) platform for 3D medical image visualization and manipulation aimed at improving the speed and accuracy of viewing and contouring on 3D images, and report our initial assessment of its performance.
Methods: We developed a software application using the Unity 3D Engine and the OpenVR API, targeting current commercially available 6-degree of freedom VR systems (e.g. the HTC Vive or Oculus Rift). This application enables 3D DICOM image and structure sets to be imported and viewed stereoscopically from arbitrary orientations. VR drawing tools are integrated to enable the creation of contours. To assess the utility and potential adoption of this technology in clinical practice, we performed an IRB-approved survey study of practicing radiation oncologists using a validated technology acceptance model.
Results: We present a new method for image review and contouring using virtual reality. A pilot study of six attending radiation oncologists (mean years of practice=19.7, range 5-23) found that 83% of participants would use this type of VR platform to augment visualization and contouring in clinical practice. On a validated questionnaire assessing new technology usefulness, the system scored a 1.8 on a 1-7 Likert scale (with 1.0 representing extremely useful).
Conclusion: Virtual reality techniques can potentially enhance a userâ€™s ability to visualize and contour on 3D medical images. Additional studies are ongoing to quantify the impact of VR on contouring accuracy and speed in performing non-planar contour expansion and assessment.